Q&A KOBI EISENBERG: Autofleet’s CEO talks taxi industry tech, cancellations, dispatch and more

Taxi companies have overseen some drastic changes over the past decade. While tackling these changes is challenging, they also provide an opportunity for growth.

We recently talked with Kobi Eisenberg, co- founder and CEO of Autofleet. Kobi is said to be a thought leader at the heart of the mobility industry, in the taxi and ride hailing space, shaping fleet operations around the world. He is also passionate about new mobility services and about providing an unrivalled customer experience that goes hand in hand with a sustainable, efficient and optimised use of resources.

The company Kobi co-founded, Autofleet, offers an optimisation platform for fleets in general and specifically taxi operators worldwide. It applies advanced machine learning algorithms to help operate taxi services at peak productivity - improving ETAs, driver availability and the number of completed rides, reducing deadhead mileage and more.

Q: How can Autofleet’s technology help taxi drivers and operators become more efficient?

I believe in the taxi industry. And I believe in the critical role taxi operators will continue to have in the value chain of mobility despite the significant challenges this past decade. But in order to grow and prosper, it is essential for minicab, PHV and taxi companies to leverage technology to automate and optimise their fleet.

Take for example automated dispatching. Many operators today dispatch to the “closest” driver based on air distance - as the crow flies. The reality however is that the actual travel time could be significantly different when considering the actual route, live traffic conditions and the direction the vehicle is moving.

These differences result in significantly higher ETAs, higher deadhead miles and overall reduced fleet efficiency. Often drivers don’t trust the dispatching decisions and take manual control that doesn’t scale. Autofleet’s Optimised Dispatching considers all of these data points in real time, across all active vehicles in the fleet, including actual route, live traffic conditions, roadblocks, vehicle driving direction, capacity constraints etc.

Once we deploy our solution, we often see an immediate improvement in the core KPIs of the operation including an improved customer experience and increased fleet efficiency. It also frees up the dispatchers from tedious tasks to deal with strategic planning and issues that require human intelligence.

We’ve combined this with machine-learning algorithms for demand prediction that accurately forecast potential demand within a 2-block radius. The prediction then generates personalised placement recommendations for drivers. We have seen an over 35% increase in completed rides through these recommendations alone. This adds up to create higher overall utilisation rates, less deadhead miles, and happier drivers that make more money.

Q: Booking cancellations are frustrating for passengers, cabbies and operators. How can Autofleet help tackle this?

The data clearly shows that the most common reason for customers cancelling the ride is high ETA, second only to constantly updating and unpredictable ETAs that change enroute to the client.

Therefore lowering cancellation rates goes hand in hand with lowering ETAs. This is possible through Optimised Dispatching and Placement Recommendation in addition to optimising the actual route for the driver.

Optimised routing calculates the ideal route across multiple parameters incorporating live traffic conditions, vehicle location data, and course heading that results in not just lower ETAs but also ETAs that are more consistent and predictable. We constantly monitor the difference between ETA and ATA (Actual time of Arrival) and usually see a difference of less than a minute

Q: What makes Autofleet uniquely different from other dispatch platforms on the market?

Autofleet has built a platform with the most advanced optimization capabilities, powered by machine learning models, that serves complicated operational needs and drives measurable business outcomes including more completed rides and better customer satisfaction scores.

But, any solution is only as good as its ability to run smoothly. Our fleet customers trust us and we take that trust very seriously, especially since dispatching is at the core of the company’s business. Which is why at Autofleet we provide more than 99.9% uptime, including 24/7 support integrated directly into the product. This means customers can chat with us and our team will respond in less than a minute - that’s our commitment!

We also feel that no one understands their market better than the taxi operators themselves. Therefore we offer operators the ability to customise their customer experience easily. For example, when we designed our passenger app, we wanted to offer more than just a logo and colour change on an app. This is why we provide our clients an open source passenger app that allows them to retain control and IP ownership over how they engage with their customers.

Finally, we also support the concept of a Universal Fleet - utilising fleet assets to generate more business, serving any source of demand. This can include delivery services, NHS rides, schools and others all on the same fleet.

Q: Now that coronavirus restrictions have largely disappeared in the UK, do you see taxi pooling making a resurgence?

If done right, pooling offers a cheaper ride for passengers and at the same time higher income for the drivers and operators - win win. However, one must be thoughtful when offering pooled taxi rides not to degrade regular service, and make sure that the passengers are not delayed for more than a set, predictable, amount of time. We designed the platform in a way that allows the operator to configure the maximum amount of time a passenger could be delayed, compared to a direct route, and the platform enforces this automatically.

Q: How can Autofleet help predict where future taxi demand might be for taxi drivers and operators?

Predicting demand and ensuring drivers are where they are needed, at the right place at the right time, is done using advanced machine learning models. These models are based on dozens of features including seasonality (hour of the day/week etc.), weather (temperature, humidity, precipitation etc.) and both planned and unplanned events (sports matches, roadblocks, strikes etc.).

We predict demand across all channels including mobile app, call centre, street hailing and partner channels, and at varying levels of time and geographic granularity depending on the specific operations. Based on the predicted demand we are able to send personalised placement recommendations to drivers, plan driver shifts and calculate prices dynamically (if applicable), all to ensure a high level of service, reliability and exceptional customer experience.

This is not a one size fits all solution, and is tailor made for each operation and market. After all, the demand for taxis in central London does not behave the same as the demand in suburban Manchester, for instance.

Q: What should taxi operators think about when they consider electrification?

ֿThere’s a big push for taxis to go electric, and this makes a lot of sense. For high utilisation use cases the total cost of ownership (TCO) of an electric vehicle is lower, as the higher initial investment is offset by energy and maintenance savings. This is of-course in addition to reducing emissions and transitioning to a sustainable future.

However, taxi firms must ensure they are not letting the customer and driver experience slip, and that the number of rides completed each day does not fall due to the downtime incurred by the charging time. With the right planning tools, taxi operators can predict the effect of switching to EVs, and work out any issues ahead of time.

Simulations in particular, can help better understand the needs of each fleet, taking into account both range and charging strategy, and help operators plan their electrification strategy at much lower risk.

Q: Some taxi operators are already using a dispatching solution. How can they assess the value of advanced optimization for their business?

I can definitely understand the concerns surrounding the transition to a new dispatching solution, and they are justified. Dispatching is at the core of any taxi business, and without it, things can grind to a halt.

However, such a critical system should really be best-in -class to ensure the business operates as well as it can. You need a state-of-the-art dispatching solution, backed by a reliable partner that supports you before, during and after deployment, with immediate response time and a deep commitment to maximise results.

To mitigate the risks involved in introducing changes to a live operation and in order to fully realise the potential of the changeover, we provide an accurate road map based on simulations, using real fleet data.

Our team of experts build a personalised plan with the team in house to facilitate the integration process and ensure the success of any new deployment.

I started this interview saying that I believe in the taxi industry, and I really do. Given the right tools, private hire, minicabs and taxi companies are well positioned to embrace and gain from new mobility opportunities that are unfolding now and in the near future.